The 2022 Midterms
in Rural America.

Standout candidates don’t
reverse long-term trends.

Donors at all funding levels, from those giving a five-dollar one-time contribution to million-dollar super PAC operations, all ask the same question: what’s the best use of a political dollar? This report from The Rural Desk is our attempt to give data-backed insights to help political donors maximize their impact, especially during a midterm election cycle. Our take: when it comes to cutting Republican margins in rural America, money is best spent on year-round partisan capacity building that stays in the district.

When it comes to cutting Republican margins in rural America, standout candidates don’t reverse long-term trends. If your aim is to reverse the current political trends, your money is best spent on year-round partisan capacity building that stays in the district. That’s our assessment based on the data we examined from the 2022 election cycle.

Democrats lost the most ground in rural America in 2022. In nearly every 2022 Senate race we could compare, the Democrat did worse in non-metro counties (RUCC 4–9) than in metro counties (RUCC 1–3), measured against the same seat in 2016. During the 2022 election cycle, the median Senate Democrat ran 1.2 points behind 2016 in non-metro counties and 2.7 points ahead in metro counties. Within the same races, with the same two candidates, the gap between metro and rural counties was more than six points, and Senate Democrats’ share of the rural two-party vote fell from 32.4% to 28.3%.

In the 2022 House races, Democrats lost ground in every kind of district but lost the most ground in the rural ones. Republicans won 75 of the 81 most rural districts, and that is where their House majority was built. Measured against the 2018 midterm instead of 2020, the rural loss grows to 5.6 points.

Nationally, the two parties’ House nominees spent about the same, but in the most rural districts Republican nominees outspent Democrats by a wide margin, and the typical rural Democratic campaign spent a small fraction of what its Republican opponent did.

When we compared districts where the same Democrat ran again with districts where a new Democrat took over, a strong showing in one election barely predicted the next, and it made little difference whether the candidate was the same person. That suggests a standout candidate’s edge doesn’t build lasting support in a district.

Our findings

In 26 of 28 Senate races,
Democrats did worse outside the metros.

In 26 of the 28 Senate races we could compare during the 2022 election cycle, the Democratic candidate did worse in non-metro counties than in metro counties, measured against the same seat in 2016.

Twenty-eight states held a regular Senate race in 2022 with a Democratic candidate facing a Republican and county results we could verify. All were for the Class III seats, last on the ballot in 2016, and together, they covered 1,802 counties. In each county, we compared the Democratic Senate candidate’s share of the two-party vote with the Democratic share in the same seat’s 2016 race. In non-metro counties, the results ranged from North Dakota’s Katrina Christiansen, who gained 8.2 points on the 2016 result for the same seat but ultimately lost her race, to New York’s Chuck Schumer, who won re-election but lost 15.9 points in his state’s non-metro counties, a spread of about 24 points.

About two in five counties (41%) ran ahead of 2016. In North Dakota and Vermont every county did, and in Ohio, New Hampshire, Iowa, Maryland, Arizona and Georgia about four in five or more did. In New York, Oregon and Hawaii no county did, and in Missouri, Arkansas and Kentucky almost none.

Only two Democrats gained more in non-metro counties than in metros

During the 2022 midterm elections, thirteen Senate Democrats gained non-metro ground compared to 2016, but only Peter Welch in Vermont and Catherine Cortez Masto in Nevada gained more there than in their states’ metros. Welch gained 6.3 points in Vermont’s non-metro counties and 5.4 in its metros. Cortez Masto gained 0.6 points in Nevada’s non-metro counties while losing 0.9 in its metros, and won by less than a point. Everywhere else the metro gain was larger: Tim Ryan gained 3.6 points in Ohio’s non-metro counties and 9.4 in its metros, and Mark Kelly gained 0.6 in non-metro counties and 10.0 in metro counties in Arizona.

Nine Democrats ran for the same seat in both years, and the rural gap held for eight of them. Maggie Hassan, Michael Bennet and Chris Van Hollen gained non-metro ground but gained more in their metros. In contrast, Chuck Schumer, Ron Wyden, Tammy Duckworth, Patty Murray and Brian Schatz lost non-metro ground, Schumer by 15.9 points, the most of the 28 Senate Democrats in the comparison, and Wyden by 9.8. Cortez Masto was the exception.

The median Senate Democrat ran 1.2 points behind 2016 in non-metro counties (RUCC 4–9) and 2.7 points ahead in metro counties (RUCC 1–3). The table below is sorted by the gap between the two. Only Cortez Masto and Welch ran better in non-metro counties than in metros, and Brian Schatz in Hawaii came close. Schumer lost the most non-metro ground but nearly as much in New York’s metros, so his gap is small. The widest gaps, about 9 points, were for Mark Kelly in Arizona, Trudy Busch Valentine in Missouri and Tammy Duckworth in Illinois.

Senate Democrats lost the most ground in rural counties

The more rural the county, the more ground Senate Democrats lost. Against the same seats in 2016, they gained 1.3 points in metro counties (RUCC 1–3) but lost 2.4 in small-town non-metro counties (RUCC 4–7) and 4.1 in rural counties (RUCC 8–9).

Comparing counties across states mixes very different races, so we also compared each county with its own state. Metro counties ran 0.8 points better than their states. Small-town counties ran 3.8 points worse, and rural counties 5.7 points worse. Within the same races, with the same two candidates, the gap between metro and rural counties was more than six points. Rural counties cast under 4% of the votes in these races, so the shift did not decide many seats on its own, but it came from a low base: Senate Democrats took 28.3% of the two-party vote in rural counties, down from 32.4% in 2016.

House Democrats lost the most ground
in the most rural districts.

In 2022, House Democrats lost ground in every kind of district, and, as in the Senate, their biggest backslide was in the rural ones. We compared each 2022 Democratic nominee’s two-party share with the Democratic share of the 2020 House vote in the same area. (Because the lines were redrawn after the 2020 Census, we re-estimated the 2020 vote on the 2022 lines, tract by tract.)

Democratic nominees in Rural and Rural-leaning districts took 35.0% of the two-party vote in 2022, down from 37.2% in 2020. A 2.1-point drop from that level is a loss of about 6% of the Democratic share, against under 2% in Urban districts (59.2%, down from 60.1%). The deepest single losses were not rural: 40 other districts fell more than six points, most of them in Florida, California and New York, against 4 of the 68 most rural.

The House map has much the same geography as the Senate map. Colorado and Pennsylvania were the states where most districts ran ahead of 2020 (6 of 8 and 11 of 13), and the typical districts in Minnesota, Ohio and Arizona also ran slightly ahead. The steepest losses came in Florida and New York, where the typical district fell 5 to 6 points.

In the map below, gray districts had no Democrat or no Republican on the 2022 ballot, or not enough contested 2020 territory to compare (81 of 435).

How we classified rural districts

We sorted all 435 districts on the 2022 lines into five tiers by the share of residents living outside Census urban areas: Urban (under 10%), Suburban (10–25%), Mixed (25–40%), Rural-leaning (40–50%) and Rural (50% and over). Measuring rurality by district rather than by county changes which seats count as rural. A district can be mostly rural even when most of its residents live in counties classified as metro, because many metro counties contain large rural areas with large rural populations.

Rural districts make up nearly a fifth of the House, and they are spread across the country rather than packed into one region. Of the 435 districts, 81 were Rural (43) or Rural-leaning (38). North Carolina had six, Georgia five, and Michigan, Mississippi, New York, Ohio, Pennsylvania and Tennessee four each.

The most rural districts lost the most ground

Republicans’ House majority was built in rural districts. Republicans won 75 of the 81 Rural and Rural-leaning districts, or 93%. Democrats won 80% of Urban districts (148 of 185), fewer than half of Suburban districts (41 of 89), 23% of Mixed districts (18 of 80) and 6 of the 81 most rural. Outside those 81 districts, Democrats won 207 of the remaining 354 seats.

Democrats lost 2.1 points in Rural and Rural-leaning districts against the 2020 House vote, compared to 0.9 to 1.8 points elsewhere: 0.9 in Urban districts, 1.8 in Suburban and 0.9 in Mixed. Sixteen of the 68 rural nominees with a 2020 benchmark ran ahead of 2020.

2022 looks even worse for Democrats when we compare the results to 2018

Every House seat was last on the ballot in 2020, so we compare each 2022 race with the last race in the same territory. However, it’s important to note that 2020 was a presidential year, with higher turnout and the president’s party not yet facing its usual midterm penalty. When we measure the 2022 election results against the 2018 midterm, we see that Democrats lost even more everywhere and the rural gap is wider: 5.6 points in Rural and Rural-leaning districts, against 2.9 to 3.6 elsewhere.

Rural Democrats were outspent
more than three to one.

Nationally speaking, during the 2022 cycle, the two parties’ House nominees spent about the same: $820 million for Democrats and $718 million for Republicans. However, there was a significant difference when it came to where the money went. In the 81 most rural districts, Republican nominees spent $146 million and Democratic nominees $44 million. The typical rural Democratic nominee spent $60,000; the typical Republican spent $1.4 million.

Democratic House spending was concentrated in urban districts and very thin in small towns and rural communities. Democratic nominees spent $462 million in Urban districts, more than twice the Republicans’ $196 million. In Rural-leaning districts they spent $14 million against the Republicans’ $64 million, and in Rural districts $30 million against $82 million. Those 81 districts received 20% of all Republican nominee spending and 5% of Democratic nominee spending.

In fact, Democratic spending per district fell from about $2.5 million in Urban districts to $370,000 in Rural-leaning districts and $710,000 in Rural ones. Republican spending per district went the other way: about $1.1 million in Urban districts and $1.7 million to $2.3 million in every other tier.

Democrats have left rural races uncontested for two decades

From 2002 to 2022, Democrats left roughly one in ten mostly nonmetro districts without a nominee, more than twice the rate in metro districts, while Republicans have almost never left a rural seat empty. In 2022, the seats Democrats left empty were almost all outside the cities and skewed rural. Of the 23 races with no Democrat, 21 were outside Urban districts, and 14 were in Mixed, Rural-leaning or Rural districts. The seats Republicans left empty were all urban. All five races with no Republican were Urban districts in Illinois, Massachusetts, New York and Pennsylvania.

So how does this compare with past cycles? Historically speaking, the empty ballot line comes in waves. In 2002 and 2004, Democrats left about one in five mostly nonmetro districts without a nominee (20.5% and 18.9%). In 2006 that fell to one district in 36, and it stayed low through 2014. In 2016 it jumped to a quarter of mostly nonmetro districts (9 of 36), the highest in the period, before falling back to a single district in 2018.

The two low points came in cycles when Democrats recruited nearly everywhere: 2006, during the national party’s 50-state strategy, and 2018, when first-time candidates filed in almost every district. Both were also strong Democratic years, so recruitment and a favorable climate arrived together. Counting nominees with little or no funding, the rural gap is wider: over the whole period, about one in six mostly nonmetro districts (16.7%) had no real Democratic campaign, against one in thirteen metro districts (7.5%).

Republicans leave House races empty too, but almost all of theirs are in metro areas. From 2002 to 2022, Republicans left 266 races without a nominee, and 223 of them (84%) were in metro districts. Democrats left 236, and most of those (138, or 58%) were in mixed or mostly nonmetro districts. Each party stays off the ballot where the other is strongest.

Year by year, each party leaves more seats open when the climate turns against it. Republicans left the most House races empty in the Democratic wave years: 45 in 2006, 41 in 2008 and 28 in 2018. Democrats left the most in the early 2000s and in Republican years: 42 in 2002, 22 in 2010 and 23 in 2022, when Republicans left just five.

In 2022, Democrats left six of the 81 most rural House districts with no candidate at all, and in 22 more their nominee spent less than $25,000. A nominee who spends under $25,000 (in 2022 dollars) is on the ballot but not really in the race. Counting those candidates, 21.2% of mostly nonmetro districts had no real Democratic campaign in 2022, up from 5.6% in 2018. In metro districts the figure was 5.1%.

In rural districts, money and
the party label set the result.

In rural districts, a Democrat’s result depends mostly on the party label and on money, and a campaign of $30,000 to $100,000 can’t move it much in either direction. A single strong result rarely repeats, and what does repeat stays with the place.

In rural America, political money and results travel together

In 2022, each tenfold increase in Democratic spending went with about 1.6 fewer points lost among rural voters.

To calculate this result, we looked at every House district with a meaningful rural vote, and estimated the Democrat’s share among rural voters. We then compared this number with the Democratic share of the 2020 House vote among the same voters.

Our analysis is that when results are lined up against what each campaign spent and how Democratic the district leaned, political spending seems to matter more in rural districts than elsewhere. Each tenfold increase in Democratic spending went with about 1.4 points more ground held in Rural and Rural-leaning districts, against about 0.7 elsewhere. That difference is not large enough to rule out chance, so treat it as a direction, not a settled finding.

Campaigns that spent under $100,000 fell 2.1 points behind the 2020 rural vote on average, and campaigns that spent over $3 million ran 0.2 points ahead of it. Holding both campaigns’ spending constant, we estimate that each tenfold increase went with about 2.4 points for Democrats and 1.6 points for Republicans.

This analysis does not prove that money causes rural gains. Our aim is simply to understand the relationship between political spending and Democratic performance.

Remember, the best-funded Democrats are more often incumbents and candidates in competitive seats, and donors give to candidates who already look strong.

Rural results leave little room for a candidate to stand out

Campaigns with more money, running in friendlier districts, usually get more votes. If you know how much each side spent, how Democratic the district usually votes and whether an incumbent is on the ballot, you can estimate about how well a Democrat should do. A Democrat who beats that estimate is what we call “running ahead of their money.” In other words, they won more votes than their budget, their district and their opponent would lead you to expect.

In rural districts, results land close to our estimate. Picture a rural Democrat who spends $30,000 in a district that voted 30% Democratic in 2020, against a Republican incumbent spending $1.5 million. Our estimate would expect this Democrat to fall about 2 points behind the 2020 result. The 2022 results typically land within about ±2.8 points of their estimate: in this case, somewhere between about 5 points behind and 1 point ahead of 2020. However, we found that in urban and suburban districts, the same kind of candidate typically lands within ±3.9 points, a range about 40% wider. Only 12% of rural Democrats beat their money by more than 2 points, against 30% elsewhere.

One good race in a rural district rarely impacts its overall trajectory

Comparing 2020 with 2022, we estimate that only about 8% of a standout result carries over to the next cycle, and more than 90% doesn’t. Most of any single overperformance reflects that year and that year alone. For this test we measured each House Democrat’s share minus Biden’s 2020 share on the same lines, again compared with what money, lean and incumbency predict, and checked whether a district’s result in 2020 predicted its result in 2022.

A new candidate in the same place tends to get about the same result. When a different Democrat ran in the same area in 2022, the correlation with the previous nominee’s 2020 result was 0.31, about the same as when the same Democrat ran again (0.29). If the candidate were driving the results on Election Day, changing the candidate should have pushed that number toward zero, and it didn’t.

Also, interestingly, we found that challengers who ran twice didn’t repeat their gains: among non-incumbents who ran in both 2020 and 2022, the correlation was effectively zero (12 candidates).

In rural districts, even less carried over. In the 34 Rural and Rural-leaning districts where a new Democrat ran in 2022, almost none of the 2020 edge returned, and among the eight rural Democrats who ran both years, Jared Golden accounts for all of the carry-over.

Why we think standout candidates don’t impact the overall trajectory of a district

For the most part, rural House races get little news coverage, and most voters never hear from the Democratic candidate because most rural Democratic candidates never raise enough money to break through the noise.

As a result, rural voters vote on the party label, so whoever the nominee is gets roughly the same partisan share. With $30,000 to $100,000, a campaign can’t reach enough voters to move that share much in either direction.

If rural results were driven by candidates, we would see a wider range and good candidates repeating their performance. Instead we see a narrow range, and gains beyond money that don’t carry over.

To move a rural result beyond what money predicts, a candidate has to become someone voters know apart from the party label, and one campaign cycle rarely does that.

The strategy for cutting rural margins
isn’t undiscovered stars.

In rural districts, “beating your money” in one race is mostly about the place a candidate runs and the year that they run, not the candidate. A place that beat its money in 2020 tended to beat it again in 2022 just as much with a different candidate as with the same one. But challengers who ran twice showed no carry-over at all. In rural districts, the only clear carry-over comes from incumbents with an established personal following.

In rural districts, the Democratic vote acts like a floor

With this analysis, we also found that most voters who will vote Democratic do so with or without a campaign because most voters vote on the party label, not the candidate.

This means that a $30,000 campaign can’t lift the Democratic vote much, but it doesn’t fall much either. In Rural-leaning districts in 2022, only three Democrats beat their money by more than 2 points: Tamie Wilson in Ohio’s 4th (+4.0), Amanda Waldman in Pennsylvania’s 9th (+2.4) and Hank Linderman in Kentucky’s 2nd (+2.1). None of them won, and none was an incumbent. Chellie Pingree in Maine’s 1st, an incumbent, was just at the line (+2.0).

Across Rural and Rural-leaning districts, eight Democrats beat their money by more than 2 points: Barry Wendell (WV-02), Tamie Wilson (OH-04), Bob Lorinser (MI-01), Naomi Andrews (OK-02), Samantha Meadows (OH-02), Amanda Waldman (PA-09), Jen Schultz (MN-08) and Hank Linderman (KY-02). They don’t share an obvious profile that would make them easy to find in advance. Compared with other rural Democrats, they spent about the same ($65,000 against $58,000), ran in redder districts (a median of 30% Democratic in 2020, against 35%) and faced more Republican money ($2.0 million against $1.3 million). None was an incumbent, and none won.

The exception is years in office

In rural America, incumbents have the biggest advantage when it comes to overperforming the Democratic brand. Jared Golden in Maine’s 2nd ran about 5 points ahead of his money in 2020 and 6 points ahead in 2022, and Chellie Pingree partly fits the same pattern. That took years in the seat, building a following that outweighs the party label, not one good campaign.

Outside group money faces the same limit

This report does not measure spending by party committees or outside groups, so what follows is an inference from the carry-over test, not a finding. Outside money is spent inside a single race, on a single candidate, in a single year. The carry-over test shows that whatever edge a Democrat gains beyond their money in one cycle mostly leaves with that year, and the same logic should apply to an edge bought by a super PAC or a party committee. An outside ad campaign can move a race while it runs, but nothing in this data suggests it leaves behind lasting support in the district. That is the same reason this report gives for not betting on standout candidates: money that arrives for one race and leaves when it’s over buys a result, not a foundation.

Invest in the place,
not the search for a star.

Democrats lost the most ground in rural America in 2022, where they were outspent more than three to one and often had no real campaign. Rural results track the party label and money closely, a strong result in one cycle rarely repeats, and the small part that does carry over belongs to the district rather than the candidate.

The data can’t say what about a district carries over. A place carries more than its partisan lean: the age, education and race of its voters, whether a local paper or TV station covers the race, how many union members live there, and whether a county party or other partisan infrastructure works there between elections. Any of these could produce the pattern.

We read the result as a case for year-round partisan capacity building that stays in the district rather than for finding standout candidates. Of the things a place carries, partisan capacity is the one Democrats can build. Demographics change slowly, and local media and union density are largely outside a campaign’s reach. If partisan capacity is what carries over, investing in it compounds.

How we’ll test it. Turning this reading into a finding takes a measure of partisan capacity. The Rural Desk’s 2026 post-election report, using methods locked before Election Day, will test whether Democrats beat expectations more where county parties are active.

How we did this

Definitions

• County types. USDA Rural-Urban Continuum Codes (RUCC) 2023: Metro (RUCC 1–3), Small Town Non-Metro (RUCC 4–7), Rural (RUCC 8–9). “Non-metro” means RUCC 4–9.
• District rurality. Share of a district’s 2020 Census residents living outside Census urban areas, summed from blocks on the 2022 district lines (U.S. Census Bureau). Five fixed tiers: Urban (<10%), Suburban (10–25%), Mixed (25–40%), Rural-leaning (40–50%), Rural (50%+).
• Change. For House races, the 2022 Democratic nominee’s share of the two-party vote minus the Democratic share of the two-party U.S. House vote in 2020 in the same territory. For Senate races, the 2022 Democratic candidate’s share minus the Democratic share in the same seat’s 2016 race, county by county. All cohort figures are weighted by 2022 votes.

The benchmarks

Senate. The 2022 regular races were for the Class III seats, so each is compared with the same seat’s 2016 race, county by county. The 2016 county votes come from MIT precinct returns; a state is used only if its county sums match MIT’s certified statewide Democratic and Republican totals within 1%. Party labels are blank or abbreviated in several states’ files, so votes are assigned by party label or by nominee surname, whichever matches the certified totals, and straight-party votes are added where the state counted them for the nominee. 2016 was a presidential year and 2022 a midterm, so the comparison also carries the difference in turnout and in the top of the ticket.

House. Every House seat was last up in 2020, so the House benchmark is the 2020 House vote in the same territory, the same last-race rule the Senate uses. Only contested 2020 races count; a county or district piece whose 2020 race had no Democrat or no Republican has no benchmark. A race counts as contested when each major party drew at least 5% of the two-party vote, so stray write-ins do not make an uncontested race look contested. The official results file marks New Jersey’s 6th and Ohio’s 7th as uncontested in 2020, but both had a Democrat and a Republican on the ballot, so both are kept. County votes come from precinct returns, calibrated to each district’s official 2020 totals so the precinct files set only how the vote splits across counties. For House districts, each 2020 county-district piece is spread over its 2020 Census tracts by population and the tracts are placed in their 2022 districts. A 2022 district is compared only if contested 2020 races cover at least 90% of its residents. Gaps: Louisiana (2020’s November jungle primary) and Alaska (no county codes in the precinct file). As a check, the same comparison was run against the 2018 midterm House vote; it shows larger losses in every tier and a wider rural gap (see the 2018 comparison at the end of the House results).

Data sources

• 2022 Senate. MIT Election Data + Science Lab precinct returns, aggregated to county and checked against certified statewide totals.
• 2022 House. Federal Election Commission, Federal Elections 2022 official results.
• 2016 Senate. MIT Election Data + Science Lab precinct returns (U.S. Senate Precinct-Level Returns 2016), aggregated to county and checked against certified statewide totals.
• 2020 House. MIT Election Data + Science Lab precinct returns; OpenElections for Indiana; calibrated to official 2020 results.
• 2018 House (check only). MIT Election Data + Science Lab precinct returns; OpenElections for New York and Maine; calibrated to official 2018 results.
• Rurality. USDA ERS RUCC 2023; U.S. Census Bureau urban/rural block population by 118th Congress district.
• House results 2002–2006. MIT Election Data + Science Lab constituency returns; district lines from Lewis et al., UCLA (108th–110th Congress). Seat counts match official totals for every year from 2002 to 2022.
• Campaign spending. FEC candidate summary files, total disbursements for the 2022 cycle.

The money analysis

Rural votes within each House district are estimated by weighting each county’s votes by its Census rural population share; the 2020 rural House share is estimated the same way from the re-estimated 2020 vote. Rural change is the difference. It is regressed on the logarithm of campaign spending and the district’s overall 2020 lean. Districts whose 2022 precinct totals did not match official results within one point were dropped (38, mostly in New York and Indiana), as were races uncontested in 2022 and districts without a 2020 benchmark (160 districts remain).

Spending by party sums total disbursements by each district’s Democratic and Republican general-election nominees’ principal campaign committees for the 2022 cycle, matched to FEC records by district and surname. It counts primary-season spending by the eventual nominee but not by primary losers, and it excludes party committees (DCCC, NRCC) and outside groups, which concentrate in competitive seats. The two-party comparison regresses Democratic rural overperformance on the logarithm of both nominees’ spending, district lean and an incumbent indicator for each party (142 districts).

The carry-over test uses every contested House race in 2020 and 2022 with campaign spending for both nominees. For each year, the Democrat’s share of the two-party vote minus Biden’s 2020 two-party share on that year’s lines is regressed on the logarithm of each nominee’s spending, Biden’s share and whether each nominee won the seat in the previous election. The residual is how far the Democrat ran ahead of or behind what those factors predict. Candidates are matched across years by state and first and last name. Each 2022 district is paired with its geographic predecessor, the 2020 district most of its residents came from (Daily Kos Elections). The figures in the text are the slopes of the 2022 residual on the 2020 residual.

The beyond-money analysis

“Beyond money” is how far a Democrat ran ahead of or behind what money, lean and incumbency predict. For every contested 2022 House district with a 2020 benchmark (354 districts), the change in the Democratic two-party share from the 2020 House vote is regressed on the logarithm of each nominee’s campaign spending (floored at $5,000), the 2020 Democratic share and an incumbent indicator for each party; the residual is the result beyond money. The model explains about 7% of the variation between districts, so the residual mostly reflects things the model cannot see. Rural and Rural-leaning districts (68) have a residual standard deviation of 2.8 points, against 3.9 in other districts (286). Within each group, each tenfold increase in Democratic spending went with 1.4 points (rural) and 0.7 points (other); the difference between the two is not statistically significant. Incumbency comes from FEC candidate records; one rural district (Minnesota’s 8th) was flagged because of a former member’s record, and no Democratic nominee among the eight rural candidates who beat their money by more than 2 points was an incumbent.

The open-seat analysis

Every House general election from 2002 to 2022 is classified by whether a Democrat and a Republican appeared on the November ballot. District rurality is the share of a district’s residents (2020 Census tracts) in nonmetro counties, placed in each year’s lines. A race has “no real campaign” when there is no Democratic nominee or the nominee’s committee spent under $25,000 in 2022 dollars. The Senate test compares counties within the same race: the Democratic Senate share minus the presidential share, regressed on the share of the county’s residents in a district with no Democratic House nominee, with race fixed effects, a nonmetro-by-year term and the county’s presidential lean. The presidential test regresses each county’s four-year presidential swing on the same open-seat share over the cycle, with state-by-year fixed effects, education, race, age, the starting level and the prior swing. Both are associations, not causal estimates.

Limitations

• No measure of partisan capacity. The report has no data on county party activity, volunteer programs or year-round field staff in these districts. The carry-over test shows that something tied to the place persists; it does not identify partisan capacity as that something.
• No demographic controls. Every comparison here is descriptive. Education, age and race are not controlled, so nothing here attributes a result to rurality itself.
• Candidate effects. Each change mixes the national environment, incumbency, candidate quality and the opponent — in both years. A Senate Democrat is measured against the 2016 nominee for the same seat, who was the same person in nine races and a different one in the rest. A House nominee is measured against the 2020 House candidates in the same territory, and some rural districts had strong Democratic candidates in 2020 who were gone by 2022. A large change is a sign of a strong or weak campaign, not a measure of one.
• The rural vote estimate in the money analysis assumes a county’s rural and urban residents vote alike. They do not; treat individual-district figures as approximate.
• Carry-over test. It covers one pair of elections, 2020 and 2022. Few rural Democrats ran in both years (eight), so the rural figures show a direction, not a settled estimate. Candidates were matched by name, which can miss a candidate who changed how their name appears on the ballot.
• Outside money. Party and super PAC spending is not counted. It concentrates in competitive seats, which are rarely rural, so the rural gap in candidate spending likely understates the overall gap.
• Open seats and selection. Democrats decide where to field candidates partly on how they expect to do, so a weak brand can cause an empty ballot line as easily as the reverse. The controls narrow this but cannot remove it.
• Coverage. Six 2022 Senate races are not in the county comparison (see the note at the end of the Senate results). Seat counts cover all 435 House districts; the change from 2020 covers the 354 of 397 contested districts with a 2020 benchmark.